Data Analysis and Knowledge Discovery  2020, Vol. 4 Issue (8): 50-62    DOI: 10.11925/infotech.2096-3467.2019.1292
 Current Issue | Archive | Adv Search |
Question Classification Based on Bidirectional GRU with Hierarchical Attention and Multi-channel Convolution
Yu Bengong1,2,Zhu Mengdi1()
1School of Management, Hefei University of Technology, Hefei 230009, China
2Key Laboratory of Process Optimization & Intelligent Decision-making, Ministry of Education, Hefei University of Technology, Hefei 230009, China
 Download: PDF (1221 KB)   HTML ( 5 )  Export: BibTeX | EndNote (RIS)
Abstract

[Objective] This paper proposes a method to extract multi-level features from the question texts, aiming to better understand their semantics and address the issues facing text classification. [Methods] First, we constructed multi-channel attention feature matrices based on the multi-feature attention mechanism at the word level. It enriched the semantic representation of the texts and fully utilized the interrogative words, properties and position features from the questions. Then, we convolved the new matrices to obtain phrase-level feature representation. Third, we rearranged the vector representation and fed data to the bidirectional GRU（Gated Recurrent Unit） to access forward and backward semantic features respectively. Finally, we applied the latent topic attention to strengthen the topic information in the bidirectional contextual features, and generated the final text vector for the classification results. [Results] The accuracy rates of proposed model with three Chinese question datasets were 93.89%, 94.47% and 94.23% respectively, which were 5.82% and 4.50% higher than those of the LSTM and CNN. [Limitations] We only examined our new model with three Chinese question corpus. [Conclusions] The proposed model fully understands the semantic features of question texts, and improves the performance of question classification.

Received: 02 December 2019      Published: 21 May 2020
 ZTFLH: TP391
Corresponding Authors: Zhu Mengdi     E-mail: 2466004852@qq.com
 The Architecture of HAMCC-BGRU αie=innerproduct(xe,xi) (1)"> Construction of Word Vector Matrix Based on Different Attention Mechanisms$αie=innerproduct(xe,xi)$ (1) Chinese Question Category System Experimental Environment and Configuration Parameter Settings Performance of Different Vector Dimensions Training Time Spent on Different Vector Dimensions The Classification Accuracy of Different Models Classification Accuracy of Different Models The Effect of Different Attention Mechanisms on the Accuracy of the Model